J. Avian Biol. 40: 625634, 2009 doi: 10.1111/j.1600-048X.2009.04621.x # 2009 The Authors. J. Compilation # 2009 J. Avian Biol. Received 13 June 2008, accepted 23 March 2009 Diurnal patterns of body mass change during stopover in a migrating songbird, the northern wheatear Oenanthe oenanthe Julia Delingat, Volker Dierschke, Heiko Schmaljohann and Franz Bairlein J. Delingat (correspondence), H. Schmaljohann and F. Bairlein, Inst. of Avian Res., An der Vogelwarte 21, 26386 Wilhelmshaven, Germany. E-mail: jdelingat@gmx.de.-V. Dierschke, Institute of Avian Research, Inselstation Helgoland, Postfach 1220, 27494 Helgoland, Germany. In migrating birds body mass change during stopover is often used to evaluate the quality of a stopover site. Because such body mass changes are difficult to survey in migrating birds various methods were developed to allow the analysis of larger sample sizes. In this article we present patterns of repeated body mass measurements of individual birds and the commonly used method of plotting body mass over time of day of birds being trapped only once. We repeatedly measured body mass of 89 northern wheatears Oenanthe oenanthe, when stopping over at a small island in the North Sea. A balance beneath bowls offering mealworms ad libitum was used to weigh their body mass several times per day. From these repeated body mass measurements we have generated a general model of daily mass gain patterns and nocturnal mass loss. Daily body mass changes followed in general an asymptotic curve progression with highest gain rates in the morning hours. During night birds lost about 5% of the evening body mass irrespectively of local wind force, temperature, night length or precipitation. By plotting first traps by time of day no such pattern in body mass increase was evident. Even in a simulated situation with repeatedly measured birds increasing their body mass, no such body mass increase could be shown when plotting one randomly chosen measurement of each bird per day. This ‘‘first-traps-by- time-of-day-method’’ depends highly on sample size, overall mass increase and the mass variation between individuals and will produce a traceable body mass increase only under certain circumstances. For a migrating bird a successful fat and protein deposition during stopover is a major requirement to reach either breeding or wintering quarters successfully and in time (Alerstam 1990). Body mass changes at stopover have a potential to indicate habitat quality, condition of individual birds, influence of infections, predation pressure or weather conditions. Furthermore, the rate of body mass change is supposed to influence departure decisions (e.g. Alerstam and Lindstro ¨m 1990, Alerstam and Hedenstro ¨m 1998). Within the past decades various studies paid special attention to body mass changes or so called fuel deposition rates in resting migrants (e.g. Carpenter et al. 1983, Lindstro ¨m and Alerstam 1992, Winker et al. 1992, Morris 1996, Wang and Moore 1997, Woodrey and Moore 1997, Fransson 1998, Schaub and Jenni 2000a, b, Da ¨nhardt and Lindstro ¨m 2001, Dunn 2001, 2002, Dierschke et al. 2005, Schmaljohann and Dierschke 2005, Bayly 2006). Fuel deposition rates are usually described as the change of body mass in gram per day or in percentage of lean body mass per day. This requires that individual birds are measured at least twice and that 24 h or a multiple of 24 h passes between those measurements. Such measurements are usually difficult to obtain in free living birds on migration, because it requires trapping and retrapping events at fixed time intervals. Methods of attracting individually marked birds to remotely controlled balances are very useful to increase the sample size of repeated body mass measurements (Carpenter et al. 1983, Lindstro ¨m et al. 1990, Fransson 1998, Da ¨nhardt and Lindstro ¨m 2001, Schmaljohann and Dierschke 2005, Bayly 2006). However, obtaining large data sets of repeated body mass measurements in certain time intervals during stop- over is time-consuming. Therefore, another method of calculating mass change at stopovers has become widely used, a method which depends on body mass measurements of birds trapped only once, (hereafter called ‘‘first-traps-by- time-of-day-method’’). Here, the pattern of body mass of first traps plotted by time of day is assumed to mirror the mean diurnal changes of body mass, which can then be calculated by linear regression analyses (Winker et al. 1992, Morris 1996, Woodrey and Moore 1997, Dunn 2000, 2001, 2002, Schaub and Jenni 2000a, Jones et al. 2002). It is often assumed that body masses increase linearly over time of day (e.g. Moore and Kerlinger 1987, Moore and Wang 1991, Winker et al. 1992, Dunn 2001, 2002, Jones et al. 2002). However, this assumption is sometimes difficult to prove, because the individual variability in body mass can be so high and thus very much concealing any possible relations (e.g. Winker et al. 1992). Regression models of first traps by time of the day were also used to correct body mass measurements for time of the day when calculating fuelling rates of birds that were 625